code logs -> 2018 -> Wed, 03 Jan 2018< code.20180102.log - code.20180104.log >
--- Log opened Wed Jan 03 00:00:21 2018
00:18 Vornotron [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Operation timed out]
01:31 celticminstrel [celticminst@Nightstar-m9434e.dsl.bell.ca] has joined #code
01:31 mode/#code [+o celticminstrel] by ChanServ
02:52 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
04:36
<@himi>
http://pythonsweetness.tumblr.com/post/169166980422/the-mysterious-case-of-the-linux-page-table
04:37
<&McMartin>
The followup post is grimly amusing as well.
04:39
<&[R]>
McM posted that two days ago
04:45
<@himi>
I missed it
04:45
<@himi>
McMartin: URL for the followup?
04:49
< Mahal>
followup is linked from original.
04:51
<@himi>
Yeah, found it
04:56 celticminstrel [celticminst@Nightstar-m9434e.dsl.bell.ca] has quit [[NS] Quit: KABOOM! It seems that I have exploded. Please wait while I reinstall the universe.]
05:00 Derakon is now known as Derakon[AFK]
05:07 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
05:07 mode/#code [+qo Vornicus Vornicus] by ChanServ
06:43 Vornlicious [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has joined #code
06:46 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
07:48 Alek [Alek@Nightstar-7or629.il.comcast.net] has quit [Ping timeout: 121 seconds]
07:52 Alek [Alek@Nightstar-7or629.il.comcast.net] has joined #code
07:52 mode/#code [+o Alek] by ChanServ
08:36 Vornlicious [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has quit [Ping timeout: 121 seconds]
08:53 Vorntastic [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has joined #code
09:05 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
09:18 Vorntastic [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has quit [Ping timeout: 121 seconds]
09:33 Vorntastic [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has joined #code
11:51 Kindamoody[zZz] is now known as Kindamoody
12:12 Kindamoody is now known as Kindamoody|afk
12:25 Namegduf [namegduf@Nightstar-88jb61.dyn.plus.net] has quit [Ping timeout: 121 seconds]
13:22
< Jessikat>
https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/
13:35 Jessikat` [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
13:35 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
13:46
<@TheWatcher>
Jessikat`: see also http://pythonsweetness.tumblr.com/post/169166980422/the-mysterious-case-of-the-linux-page-table from McM a couple of days ago. This one's going to be a real clusterfuck >.<
13:48
<&McMartin>
Embargo apparently is lifted tomorrow
13:48
<&McMartin>
Should be fun times
13:48
<&McMartin>
(The Register article is better sourced)
14:03
< Vorntastic>
This is the crazy baddest
14:03
< Jessikat`>
Heh
14:17 Jessikat` [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Connection closed]
14:17 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has joined #code
14:18
<@TheWatcher>
It could be worse
14:18
<@TheWatcher>
It could set us on fire.
14:32 Namegduf [namegduf@Nightstar-08u.5uh.185.31.IP] has joined #code
14:32 mode/#code [+o Namegduf] by ChanServ
14:46
< Vorntastic>
That is true, it could be an hcf
15:02 Jessikat [Jessikat@Nightstar-our0up.cpe.teksavvy.com] has quit [Ping timeout: 121 seconds]
15:21 Degi [Degi@Nightstar-88rhip.dyn.telefonica.de] has joined #code
15:24
<&[R]>
So glad I'm in bed with AMD now
15:28
<@TheWatcher>
I'd second that, except I'd half fear some cockup from AMD coming to light~
15:28
<&[R]>
Aye
15:29
<&[R]>
All my Intel CPUs are pretty slow already TBH
15:29
<&[R]>
So this isn't going to be a fun patch to apply
15:30
<&[R]>
(I think I have 5/19)
15:34
< Vorntastic>
Saw a thing, might have been here, where it was mentioned that if it goes back as far as described there hasn't been a bug-free Intel processor since the 486
15:35
< Vorntastic>
What with f00f and the division bug and some other fool thing that I don't remember
15:38
<@Tamber>
Hooray complexity~
15:38
<@Tamber>
(Delicious, delicious complexity...)
16:07 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
16:07 mode/#code [+qo Vornicus Vornicus] by ChanServ
16:20 Kindamoody|afk is now known as Kindamoody
17:14 Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code
17:23
<&[R]>
https://cyber.wtf/2017/07/28/negative-result-reading-kernel-memory-from-user-mode/ <-- more of that thing
17:28
<~Vornicus>
the block diagram there finally explains to me a difference between the various levels of cache
17:39
<&[R]>
Always nice to learn something
17:40
<&[R]>
It boggles my mind that someone could attack the CPU itself
17:40
<&[R]>
I mean, I started to get that impression when they started attacking IME
17:40
<&[R]>
But this is a whole other level of crazy
17:41
<&[R]>
(The IME attack in question was locally run code that did /not/ hit the network.)
17:47
<&[R]>
<numbdewd> this guy looks sly too,, doesn't he :P https://www.fool.com/investing/2017/12/19/intels-ceo-just-sold-a-lot-of-stock.aspx .. (frmo #Openbsd just now - thatlink i mean) <-- also that
18:20
<@Tamber>
https://twitter.com/brainsmoke/status/948561799875502080
19:25
< Vorntastic>
Hot dang
20:28 Vorntastic [Vorn@Nightstar-vn465g.sub-174-210-8.myvzw.com] has quit [[NS] Quit: Bye]
20:28 Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
20:28 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds]
20:29
<&McMartin>
Sometimes people wonder why I'd retrocode in assembly language for fun
20:29
<&McMartin>
Some of the answer is days like today, where one does in fact recognize the value in a CPU made of only 4,000 transistors.
20:32
<&McMartin>
Well, OK. 5000. But 1000 of them are functionally just resistors.
20:38
<@gnolam>
:)
20:42
<&[R]>
https://newsroom.intel.com/news/intel-responds-to-security-research-findings/
20:43
<&[R]>
"we haven't fucked up, and our CEO selling stock means absolutely nothing. Move along, move along."
20:43
<@TheWatcher>
"Intel believes its products are the most secure in the world" It's good to have beliefs. They say it's better for your mental health.
20:44
<&[R]>
Ha
20:44
<@TheWatcher>
Also, I'm glad it's not a "bug" or a "flaw", but is instead an "issue".
20:44
<@TheWatcher>
Comforting, that.
20:45
<&McMartin>
To be fair, barely half the things on my "buglist" are actually classified as "bugs"
20:45
<&McMartin>
https://www.reddit.com/r/pcmasterrace/comments/7nthay/a_quick_summary_of_the_current_intel_cpu_bug/
20:47
<&[R]>
"The issue impacts all modern Intel CPUs. (Edit: It's been confirmed that the latest unaffected CPU is the original Pentium.)"
20:47
<&McMartin>
I'm seeing claims that it allows system takeover but so far all actual explanations of how it works only let you *read* kernel memory
20:47
<&McMartin>
Which lets you defeat ASLR if you have already otherwise compromised the system
20:48
<@Tamber>
one part of a chain of exploits
20:49
<&McMartin>
Yeah, unless terminology has changed since I was actually explicitly learning stuff, you don't name the severity of the exploit by the severity of the most powerful chain of attacks that contains it, you name the chain that
20:49
<&[R]>
The chain that what?
20:50
<&McMartin>
"that" is a noun there, not a preposition
20:50
<&McMartin>
If you have an exploit that defeats ALSR, and another exploit that is a privilege escalation
20:50
<&McMartin>
And then you have an attack that uses both
20:51
<&McMartin>
Then ALSR defeats #2, is defeated by #1, but an ALSR-defended system is only "taken over" by #3.
20:52
<&McMartin>
In particular, that "negative result" link a few screens back seems to tell us that you can't *write* kernel memory from user code with this.
22:11
<&[R]>
https://github.com/xoreaxeaxeax/movfuscator
22:11 Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has quit [Ping timeout: 121 seconds]
22:30
<&McMartin>
I suppose xor eax, eax, eax is a slightly obfuscated mov
22:53 Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code
22:53 mode/#code [+qo Vornicus Vornicus] by ChanServ
23:10
<&McMartin>
Hey, here we go. https://googleprojectzero.blogspot.com/2018/01/reading-privileged-memory-with-side.html
23:13 himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds]
23:19
<&[R]>
https://meltdownattack.com/
23:27
<&McMartin>
TheWatcher: re: "Issue" as opposed to "bug" or "flaw", ARM has an update out about this too that seems to suggest those are terms of art:
23:27
<&McMartin>
"However, this side-channel mechanism could enable someone to potentially extract some information that otherwise would not be accessible to software from processors that are performing as designed and not based on a flaw or bug."
23:29
<&[R]>
Almost sounds like "you discovered out 'acceptable tradeoff'"
23:29
<&[R]>
our*
23:30
<@TheWatcher>
Well. Next week, I could ask Steve Furber in person if that is the case~
23:30
<&McMartin>
I can imagine 'flaw' might be restricted to something like 'manufacturing defect', which this clearly isn't~
23:31
<&[R]>
The thing about name dropping, is the person listening has to know who that is :p
23:31
<@TheWatcher>
Principle designer of the ARM processor
23:31
<@TheWatcher>
*Principal
23:31
<&[R]>
Nifty
23:31
<@gnolam>
Aww. You got there before I could suggest "the inventor of the Furby".
23:32 * [R] has 3 RPi 3s now
23:32
<@TheWatcher>
Snrk
23:32
<@TheWatcher>
That said, right about now I'm considering setting up as a furniture maker, or a watch maker, at this point. Less fucking craziness.
23:33
<&McMartin>
He's probably not the one at fault for RISC OS (and maybe Arthur's?) sprite palette layout
23:33
<&McMartin>
I want to ask *someone* about that.
23:33
<&McMartin>
The one where it's got noncontiguous color bits.
23:34
<&[R]>
IE colors were sorted by someone on LSD?
23:34
<&McMartin>
Not really
23:34
<&McMartin>
BBC BASIC resorted them to make some sense.
23:35
<&McMartin>
It had, broadly, 12-bit RGB color, but only 256 colors usable
23:35 Degi_ [Degi@Nightstar-fp79ro.dyn.telefonica.de] has joined #code
23:35
<&McMartin>
So the rule was that the two lowest bits in RGB had to be the same (call it "black")
23:35
<&McMartin>
So you had COL which took a 0-63 value RRGGBB, more or less, and then a TINT that was the KK
23:35
<&McMartin>
That serialized something like RRGGBBKK, but only when doing BASIC-y things.
23:36
<&McMartin>
For the actual OS/GUI-level bitmap objects (aka, filetype "Sprite", used as icons and in-application graphics), it was... not
23:36 * McMartin goes to dig that up again
23:37
<&McMartin>
OK, right, sorry, it was BBGGRR for the BASIC-y stuff, and also for the "draw graphics primitives" syscalls
23:37
<&McMartin>
But for the sprite bitmaps, it was BGGRBRKK.
23:38 Degi [Degi@Nightstar-88rhip.dyn.telefonica.de] has quit [Ping timeout: 121 seconds]
23:38
<&McMartin>
(Still looks fine in the end, mind you. https://bumbershootsoft.files.wordpress.com/2017/07/riscos_17_target.png )
23:44
<&McMartin>
The old RiscPC display didn't quite match what DOS VGA could, though, so the plasma effects look bad in RPCEmu
23:45
<&McMartin>
RISC OS Pi uses 24-bit color and handles it fine across the board.
--- Log closed Thu Jan 04 00:00:22 2018
code logs -> 2018 -> Wed, 03 Jan 2018< code.20180102.log - code.20180104.log >

[ Latest log file ]